AI-generated research brief — verify at source
Bayesian Precision Dosing Significantly Improves Vancomycin Target Attainment in Critically Ill Children
In a multicentre randomised controlled trial, model-informed precision dosing (MIPD) of vancomycin achieved the pharmacokinetic/pharmacodynamic target in nearly 72% of critically ill children, compared with 54% under standard therapeutic drug monitoring — an absolute improvement of 18.9 percentage points (95% CI 1.7 to 34.7). The finding is statistically significant and supports the clinical adoption of Bayesian-guided dosing in this vulnerable paediatric population, though the accompanying reduction in kidney injury did not reach significance.
What Was Studied
The BENEFICIAL trial examined whether MIPD using Bayesian software and early AUC-guided sampling could outperform conventional therapeutic drug monitoring (TDM) in achieving the optimal vancomycin exposure target — a 24-hour area-under-the-curve to minimum inhibitory concentration (AUC:MIC) ratio of 400–600 mg·h/L — in severely ill neonates and children. Vancomycin’s narrow therapeutic window creates a persistent clinical dilemma: underdosing risks treatment failure, while overdosing is associated with acute kidney injury.
How It Was Studied
This pragmatic, individually randomised, superiority trial enrolled 314 critically ill patients under 18 years of age — ranging from neonates to adolescents — who were initiating intravenous vancomycin for suspected or confirmed Gram-positive infection across 14 paediatric or neonatal intensive care and haemato-oncology units in seven Belgian hospitals. Patients were assigned 1:1 to either MIPD (using Bayesian software with additional early blood sampling to estimate real-time AUC) or standard-of-care TDM. Treating physicians and pharmacists were not masked to treatment allocation, though patients, families, and the biostatistician were. Key exclusion criteria included extracorporeal support, severe or chronic kidney disease, and imminent death. The primary analysis was conducted on an intention-to-treat basis.
What Was Observed
- Target AUC:MIC attainment was substantially higher with MIPD: 71.8% of patients in the MIPD group reached the therapeutic target at 24–48 hours, compared with 53.9% in the standard-of-care group — an absolute difference of 18.9 percentage points (95% CI 1.7 to 34.7), indicating a statistically significant and clinically meaningful improvement in precision dosing.
- Acute kidney injury and mortality showed a numerical but non-significant reduction: The composite outcome of new or worsening acute kidney injury or all-cause death occurred in 12.4% of MIPD patients versus 16.9% under standard care, an absolute difference of -4.5 percentage points (95% CI -11.6 to 3.5). The confidence interval spans both a moderate benefit and a small harm, meaning this difference could plausibly be due to chance.
- Serious adverse events were equally distributed between groups: Eight patients (5%) in each arm experienced serious adverse events. One death in the MIPD group was considered at least possibly related to the vancomycin administration method, indicating that MIPD is not without risk, even if the overall safety profile appeared comparable.
Why This Matters
Vancomycin is a cornerstone treatment for serious Gram-positive infections in critically ill children, yet dosing it accurately in this population is especially challenging due to highly variable pharmacokinetics. This trial provides evidence from a large, prospective, multicentre RCT that MIPD can meaningfully close the gap between intended and achieved drug exposure. The authors note that elevated vancomycin AUC is a well-established driver of kidney injury, and the lower cumulative exposure observed with MIPD — even in the absence of a statistically significant AKI benefit — lends biological plausibility to its use in patients at greatest risk.
How to Read This Result
While the primary pharmacokinetic outcome clearly favoured MIPD in this well-designed multicentre RCT, the absence of a statistically significant reduction in acute kidney injury or mortality means that the direct clinical safety benefit remains uncertain, and the results should be interpreted with particular caution in settings that differ from the Belgian paediatric intensive care context studied here.
Limitations
The trial’s open-label design — necessary given the nature of the intervention — means that treating physicians and pharmacists knew which dosing approach was being used, introducing potential for performance bias. The pragmatic design across multiple centres and ward types may have added variability in how standard-of-care TDM was conducted. The key safety composite outcome of acute kidney injury or all-cause mortality did not reach statistical significance, leaving the nephroprotective hypothesis unconfirmed. Additionally, 18 patients were excluded post-randomisation due to deferred consent not being formalised, which introduces a minor integrity concern for the ITT analysis.